| 摘要 |
An apparatus for encoding an input digital signal divides the input digital signal to produce frequency components of increasing bandwidth as the frequency increases, performs blocking on each of the frequency components to produce blocks of decreasing time length as the frequency increases, performs a fast Fourier transformation on each block and quantizes each block using a quantization characteristic set in accordance with a comparison of the energy of each frequency component in each block with a masking spectrum. The masking spectrum includes both a masking effect on the frequency base of the input signal and masking effect on the time base of the input signal and may further include effects of the minimum audible curve of the human hearing sense characteristic. Utilization of both the temporal and frequency masking effects permits reduction in the bit rate of the encoded signal using real time processing and without deterioration in sound quality.
|
